The Model ST-4 Star Tracker / Imaging Camera is a multipurpose instrument. It can be used as
an automatic star tracker to take long guided exposures of the night sky, or, in conjunction with
a personal computer (PC), as a highly sensitive imaging camera. This manual describes the
technical concept of the Model ST-4, the interfaces to the telescope and computer, and general
operating instructions.
Tracking Overview
The Model ST-4 uses a charge coupled device
(CCD) detector to detect star images. The
detector can be understood by examining Figure
1. An array of 192 by 165 detector elements,
called pixels, is arranged as shown in the figure.
We refer to the horizontal axis as X, and the
vertical axis as Y.
X
Y
165
Pixels
192
Pixels
Enlarged View
Showing Pixels
Figure 1CCD Configuration
In operation the pixels convert photons into
electrons, and store them until read out by the
ST-4's microcontroller (the ST-4 has memory for
two images, one for light frames, and one for
dark frames, as explained further below). For
example, if a star's image is present at pixel
position 121,132 the computer will determine its
position by noting the increased signal from that
pixel. If the star is drifting due to guiding errors,
it will appear at a different position in the next
exposure, say at pixel position 123,132. The
computer then calculates how far the star has
drifted from the original exposure, and toggles
the telescope drive accordingly.
The microcontroller (in the ST-4) can take an
exposure (called integration), read out all the
pixel values, and calculate the necessary
telescope correction in less than a second. The
tremendous sensitivity of the CCD enables guide
stars as faint as 8th magnitude to be tracked with
a 1 second exposure and a 60 mm guide
telescope.
The calculating power of the ST-4 enables the
star's location to be determined to a fraction of a
pixel accuracy, enabling very accurate tracking.
Imaging Overview
The Microcontroller referred to in the tracking
section is built into the ST-4. This microcontroller
can communicate to an XT or AT compatible PC,
or a Macintosh* , over the ST-4's RS-232 serial
link (the serial interface of the ST-4 is also
compatible with the RS-422 ports used on the
Macintosh).
A full image can be transmitted at 19.2K
baud within 18 seconds and over a distance of
100 feet. Data transfer rates as high as 57.6K
*
Macintosh is a registered trademark of Apple
Computer, Inc.
SBIG ST-4/0490Page 1
SBIGSANTA BARBARA INSTRUMENT GROUP
baud are supported, and work very well over
shorter cable runs. A partial image transfer
mode is also supported which sends 1/4 as many
pixels (one value for each group of four adjacent
CCD pixels) in 5 seconds. This mode of
operation is very handy for focussing the
telescope and finding objects. Finding faint
objects is easy using this method; the outline of
the Ring Nebula is clearly seen in exposures as
short as 10 seconds with an 8" Schmidt
Cassegrain Telescope (SCT) operating at f/10.
In imaging mode, the microcontroller in the
ST-4 is told to take an exposure by the PC. It
does so, and stores the resulting data in memory
within the instrument. This data is then relayed
over the serial link to the host computer. The
data is retained in the ST-4 until the next
exposure is captured or power is turned off. The
host does all of the extensive data manipulation
required by the user, such as contrast
enhancement, and also can store data on disk for
later study. This is a very attractive feature; the
computer allows a quick look to be taken at the
image, within seconds of the event, and a more
detailed look at any later time, such as a rainy
night or during the day.
The CCD can be exposed for integration
times longer than five minutes. The pixels slowly
fill-up in the dark due to a phenomena called
dark current, and they saturate at about the five
minute point when the CCD is cooled to a
temperature near -30 °C using the single stage
thermoelectric cooler used in the ST-4. The waste
heat of the thermoelectric cooler (about two
watts) is dissipated into the air by convection
around the CCD head. The CCD has a sensitivity
comparable to ASA 20,000 film, if such a film
speed were available. The CCD has a limited
resolution due to the small number of pixels;
much greater resolution would be degraded by
the limitations of most computer graphics
screens.
Note:
With the CCD running at lower than ambient
temperatures, you will wonder why dew and
frosting aren't a problem. First of all, the
chamber containing the CCD is small, and
only a small volume of air surrounds the
CCD. The small volume minimizes the total
amount of water vapor in the air, which will
frost onto the coldest surface inside the head
(which is the bottom of the CCD). Although
frost may initially form on the top of the
CCD, in a matter of minutes it will migrate
and be trapped at the back of the CCD.
System Interfaces
The following equipment is a prerequisite to
running the ST-4 CCD Star Tracker / Imaging
Camera.
1.A Telescope with pushbutton or joystick
slow motions in at least Right Ascension
and and hopefully also Declination.
2.A guide telescope, 50mm aperture or
larger, or an off axis guider arrangement.
For pushbutton type controllers the ST-4 tracker
controls the telescope the same way you do;
through the RA and Declination slow-motion
adjustment switches, the interface to which is
shown in Figure 2. Four relays in the ST-4 are
used to operate the switches. Most telescope
drives have two Declination motor adjustment
switches which are normally in the 'open'
position. Pushing the button or closing the relay
both apply voltage to the motor. Study Figure 2
carefully, along with your pushbutton control, to
determine the correct configuration.
Most telescopes have one Right Ascension
switch which is normally closed; opening this
switch slows down the drive. The other Right
Ascension switch is normally open; closing this
switch speeds up the drive. It is apparent that
the relay contacts which are brought out on the
cable, 4 groups of three (normally open, normally
closed, and common) are all that is necessary to
control the telescope. The cable pinouts are
described in Appendix A.
When the ST-4 tracker is connected to the
telescope, the hand controller is not disabled, and
Page 2SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
still operates normally. If the telescope control is
modified for the ST-4, and the ST-4 is unplugged,
the drive may not run since the normally closed
connection in the ST-4 has now been removed. If
this situation is a problem it is best to build up a
mating connector to replace the ST-4 box that has
the appropriate pins shorted together (usually
just two are required to enable the RA drive to
work).
For joystick type controllers, the four relays
in the ST-4 are used to simulate the joystick being
pushed to its limits. Two relays are used for each
axis or rheostat of the joystick as shown in Figure
3. For the Right Ascension rheostat you would
use the +X and -X relays, and for the Declination
rheostat you would use the +Y and -Y rheostat.
A: Standard Hand Controller Switch
c
switch
common
A
rheostat
B
A: Standard Joystick
Figure 3 Joystick Interface
C
A
relay
c
nc
no
relay
nccno
BC
B: Modified Joystick
nonc
B: Modified Hand Controller Switch
c
nonc
Figure 2 Pushbutton Interface
normally open
normally closed
c
relay
nonc
common
normally open
normally closed
SBIG ST-4/0489Page 3
SBIGSANTA BARBARA INSTRUMENT GROUP
STAR TRACKING OPERATION
The instrument panel is illustrated in Figure 4.
The ST-4 instrument is furnished with a power
cable, a cable to the CCD, and a third cable for
the telescope's hand controller. There is no
power switch; plugging in the power turns the
instrument on. The instrument will come-up in
FIND AND FOCUS mode, which displays the
greatest pixel value found in the readout of the
CCD array, and the location of that pixel. The
greatest pixel reading will drop as the CCD cools
in temperature. After about 2 minutes, the CCD
will have cooled to its optimum temperature.
Note:
The display values range from 0 to 99
corresponding to percentage. Pixel X and Y
values of 50 correspond to a pixel centered
within the CCD, while a brightness value of
99 corresponds to a completely saturated
pixel, i.e., the star is too bright for tracking.
Instrument Startup for Tracking Purposes
1. Insert the CCD head into the eyepiece tube
such that it seats accurately against the tube.
The CCD will not be damaged by light when
powered, so it can be easily handled.
2. Begin calibrating the CCD by completely
blocking the open end of the telescope to
remove all light (the CCD is very sensitive it will saturate in very low ambient light
levels).
3. Observe the VALUE reading. This reading
should fall to a reading of around 10 and
stabilize within two minutes of the
instrument being powered-on. Press the
INTERRUPT button to halt the FIND AND
FOCUS mode (or any other mode for that
matter).
4.Push the TAKE DARK FRAME button. The
microcontroller will readout the CCD and put
the data in the dark frame memory.
ASTRONOMICAL
SBIG
STAR TRACKER/IMAGING CAMERA
TAKE
DARK
FRAME
FIND
AND
FOCUS
CALIBRATE
DRIVE
TRACK
Figure 4 Instrument Panel
5. Push the FIND AND FOCUS button and
uncover the telescope to begin the collection
of light by the ST-4. The instrument will
automatically begin taking frames of CCD
data, subtract the dark frame stored in
memory, and display the maximum value.
Direct the telescope to a star and adjust the
telescope's position to approximately center
INSTRUMENTS
MODESELECT
894537
VALUEXY
INTERRUPT
MENU
ADJUST
the star image on the CCD by noting the
reading on the X and Y displays.
6. If the star is too bright (VALUE reading 99)
then either the exposure must be reduced, or
a fainter star chosen for tracking. In order to
correct this condition, press the INTERRUPT
button to stop the collection of data, and
Page 4SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
return control of the ST-4 instrument to the
keyboard.
Push the MENU button. The brightness
display will then read "EA" (exposure adjust)
and a "1" will appear on the X pixel display
(indicating a default exposure time of one
second).
Repeatedly pushing the ADJUST key will
scroll through a list of exposure times, from
0.1 second, to 20 seconds. Adjusting the
exposure time to shorter times will reduce the
star brightness, while choosing a longer
exposure time will increase star brightness.
When the desired exposure time has been
chosen, press the MENU button again (the
ST-4 will display "CA" (Calibration Adjust),
and then press it several more times until you
see a bA displayed in the value box. bA
stands for brightness adjust; two modes are
available, A for average, or F for faint. F
increases sensitivity by 9 times. You should
set this parameter to A for initial
familiarization. Press the MENU button
again until you see the boost (b) parameter
displayed. This is a boost factor, where
greater values mean greater gain. Initially set
this value to 1. Press the MENU button again
to return to the normal operating mode (after
being interrupted "HELLO" appears on the
display).
Note:
You must take a new dark frame if you
change the exposure time, the brightness
adjust, or the boost factor.
Table 1. This assumes a 1 second exposure
and a typical response of the CCD. Use this
as a reference for determining whether the
system is properly focussed.
StarST-4 VALUE Reading
Magnitude60mm Refractor8 inch SCT
49999
56099
62499
79* 80
84
91
10-5
Table 1 Typical VALUE readings
8.When the focus is adjusted, remove the CCD
and insert an eyepiece into the tube, sliding
the eyepiece in the tube until the image is in
sharp visual focus. With a knife or other
sharp object, scribe the eyepiece on its side at
the end of the tube. This eyepiece can then be
used to quickly center and focus the CCD in
the future Place the CCD back into the tube
so that it seats against the tube as before.
9.Position the star image approximately in the
center of the CCD (X=50, Y=50) using the
telescope controls. Push the UP, DOWN,
LEFT, and RIGHT buttons for a few seconds
to make sure that each relay control is
correctly interfaced to the telescope handheld controller unit. If the buttons are
working correctly, the star image will move
in four different directions (but not
necessarily up, down, left, and right).
*
*
32
13
*
7.When the brightness level has been adjusted
to an acceptable level, focus the telescope by
turning the focus knob and observing the
VALUE display. At best focus, this number
is maximum. Be careful to take your hand off
the telescope between adjustments or the
telescope vibration will smear the star image
over multiple pixels within the CCD,
reducing the brightness. Atmospheric
turbulence will also tend to smear the image,
so it may be helpful to watch several
sequential exposures when critically
focussing the image.
A table of typical VALUE readings for
different magnitude stars is shown below in
SBIG ST-4/0489Page 5
10. Push the CALIBRATE button. The ST-4 will
automatically drive the telescope in each
direction, determine which direction
corresponds to +X and +Y, and calculate the
correction speed of your drive in all four
directions (in pixels moved per second). This
process takes about 30 seconds. This is a
four-step process with the ST-4 exercising the
four relays, and after each step the ST-4 will
momentarily display the location and
brightness of the brightest object in the field
of view. If the image moves too little (less
*
Use an increased boost factor and the faint star
mode when working at these levels.
SBIGSANTA BARBARA INSTRUMENT GROUP
than 2 units) or too far (greater than 30 units)
the relay closure time should be adjusted by
pressing the MENU key twice to get to the
"CA" option, and then repeatedly pressing
the ADJUST key to scroll through the time
adjustment values (1 to 20 seconds).
11. If no errors are displayed at the end of the
CALIBRATE procedure (E1 or E2 in the
Value display), then position the guide star
on the center of the CCD, adjust the exposure
time if necessary, and then press the TRACK
button. The ST-4 will automatically guide the
telescope until the INTERRUPT button is
pressed.
Note:
The ST-4 constantly corrects drive
adjustment times to try to improve tracking.
The visual X and Y displays show the
tracking error seen during each exposure in
units of 0.2 pixels (i.e. a displayed Y value of
-3 indicates that the star moved 0.6 pixels in
the -Y direction during the exposure). The
number displayed after the A in the value
location is the average error for the last 16
correction periods.
If the star is completely lost during an exposure,
the unit will stop tracking and the display will
display the brightness and location of the
brightest object in the field of view (like the FIND
AND FOCUS mode), and the track lost relay will
be activated for one second after 5 consecutive
misses. The telescope will not be corrected again
until the star reappears.
Calibration steps 10 and 11 should be
repeated whenever the telescope is substantially
re-positioned in Declination (to correct for longer
RA adjustments near the poles). Calibration
should also be repeated if the telescope mount is
of German Equatorial design and the telescope
tube flips from one side of the mount to the other
(reversing adjustment directions). Care should
be taken to orient the CCD head so the RA and
Declination axes line-up with the sides of the
CCD head (you can tell by looking through the
glass window at the CCD or by noting the
orientation of the Serial Number tag on the rear
of the head which is oriented like the CCD). This
adjustment is not critical. The tracker will work
in any orientation. It just makes it easier to make
sense out of the x and y readings and to use the
pushbuttons if the axes are lined up to the RA
and Declination axes.
Explanation of Menu Items
The menu choices which can be adjusted to a
particular telescope's configuration can be
viewed by pressing the MENU button
repeatedly. The different menu items have the
following effect:
Note:
For each menu item, pressing the ADJUST
button repeatedly will scroll through the
choices, finally jumping back to the lowest
value choice. There is no way to back up.
EA: Exposure Adjust
The exposure (integration time) used by the
ST-4 in the tracking mode can be set to be
from 0.1 to 20 seconds. The readout time of
the array is 0.14 seconds; the smearing
produced by the readout time for short
integration times is not significant in the
tracking mode.
CA: Calibration Adjust
The amount of time (in seconds) the drive is
left on during each move when the
CALIBRATE mode is executed. If the time is
too short the move will not be accurately
determined. If it is too long the star will
move off the array. Set this parameter such
that a move of 5 to 30 units result.
SA: Scintillation Adjust
The ST-4 modifies the correction factors
determined in the CALIBRATE mode, if
necessary, to improve the tracking. This
modification is performed only if errors
greater than the SA factor are seen (the SA
factor is in pixels). Telescopes with extremely
long focal lengths (>10 feet) may find the
tracking is improved if this value is increased.
Also, its setting should be increased if the ST4 shows any tendency to "run away" during
tracking.
Page 6SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
bA: Brightness Adjustment
Setting this to A is the normal mode. If this
parameter is set to F the ST-4 will set each
pixel equal to the sum of the 3 x 3 pixel box
centered on that pixel prior to performing the
tracking calculation. Since star images are
often smeared over several pixels, this
collects more light from a faint star. A new
dark frame should be taken after altering this
value.
H1: Hysteresis (backlash) Adjustment, X-axis
This parameter sets the amount of extra time
that the drive is operated when the direction
of adjustment is changed. Many telescopes
have severe backlash, and the star image may
not move for several seconds after a
correction is made. The parameter is the
number of tenth second increments added to
the calculated move when the move direction
is reversed from the previous move.
H2: Hysteresis Adjustment, Y-axis
This parameter is identical to the H1
parameter above except for the Y-axis.
b: Boost factor
This parameter boosts the internal gain of the
ST-4, enabling fainter stars to be tracked with
short exposures. Care must be taken that
thermal variations of the CCD head do not
cause the star to be lost when working with
very faint stars. If this value is altered a new
dark frame must be captured.
Please review the problem section at the end of
this manual if problems are encountered during
tracking.
SBIG ST-4/0489Page 7
SBIGSANTA BARBARA INSTRUMENT GROUP
IMAGING CAMERA OPERATION
The ST-4 works quite well as an electronic
imaging camera when connected to either an IBM
PC or compatible or an Apple Macintosh. Some
details concerning the operation of the CCD in
this mode are necessary to understand all the
controls and features available in this mode.
CCD Pixel Dimensions
The CCD pixels are not square. Their dimensions
are 0.01375 millimeters wide in the X-direction
and 0.016 millimeters tall in the Y-direction.
Note that the CCD active area is therefore 2.64 by
2.64 millimeters, or exactly square (about one
tenth inch on a side).
Caveats of CCD Readout
The CCD is read out electronically by shifting
each row of pixels into a readout register at the
Y=0 position of the CCD, and then shifting the
row out through an amplifier at the X=0 position.
The entire array shifts down one row when a row
is shifted into the readout register, and a blank
row is inserted at the Y=165 position. Note that
the CCD elements are still collecting light as they
step down to the readout register. Most
commercial CCD cameras use a more expensive
CCD which has what is known as a frame
transfer readout mode, where all active pixels are
shifted very quickly into a pixel array screened
from the light by a metal layer, and then read out
slowly. The SBIG ST-4 CCD minimizes the effect
of not having a frame transfer buffer by reading
out the array very quickly, in about 0.14 seconds.
As long as the CCD exposure is greater than
about one second this technique will reduce
streaking of the stars to acceptable levels.
Planets pose a particular problem to the CCD
since they are so bright that exposures of 1
second at f/10 are badly overexposed. The ST-4
has a Half Frame mode for planets and bright
stars to solve this problem. In the Half Frame
mode the upper half of the CCD is used as a
frame buffer for a bright image positioned in the
lower half of the CCD. A short exposure can be
taken (down to 0.01 seconds) and the bottom
half of the array shifted rapidly up to the upper
half. The 82 lines of short exposure data can then
be readout at the normal rate. This method
works quite well, and uses enough pixels such
that 0.5 arcsecond per pixel scale factors can be
achieved while viewing an entire planet.
Unfortunately, this technique does not work
for the moon, since the moon's image typically
fills the CCD. The only way to image the moon is
to use neutral density filters to attenuate the light
down to where the CCD can be used for a 1
second exposure without saturating. The same
holds true for images of terrestrial scenes during
daylight.
When a long exposure is taken, a glow will be
noticed in the upper left corner of the image, near
pixel (1,1). This is apparently due to heating of
the array by the readout electronics increasing
the dark current. This glow can saturate the
array in the corner in exposures several minutes
long and cause a blank region to appear in the
subtracted data. Using exposures short enough
that saturation does not occur in the corner will
reduce this cosmetic problem to acceptable levels.
Dark Current
The CCD can not take an unlimited exposure.
During an exposure in the dark the pixels will
slowly fill up due to an effect known as the dark
current of the device. This dark current is
reduced by cooling, and this is why the CCD is
mounted on a single stage thermoelectric cooler
in the CCD head. The window is used over the
CCD to keep moisture from the CCD while it is
powered. If the window is removed the CCD
will frost rapidly (in seconds). In a room
temperature environment the cooled CCD pixels
take about 5 minutes to fill up due to dark
current. At lower ambient temperatures the CCD
will take longer (maximum exposure time
doubles for roughly every 20 °F drop in
temperature) until around 32 °F, where the cooler
power is reduced to avoid damaging the CCD
with excessive cold. The dark current can be
subtracted from images as described in a later
section, but the effects of the dark current filling
the CCD can not be avoided.
CCD vs Film
The CCD is very good at the most difficult
astronomical imaging problem; imaging small,
Page 8SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
faint objects. For such scenes long exposures are
typically required. The CCD based system has
several advantages over film; greater speed,
quantitative accuracy, ability to increase contrast
and subtract sky background with a few
keystrokes, the ability to co-add multiple images
to increase sensitivity without tedious darkroom
operations, wider spectral range, and instant
examination of the images at the telescope for
quality. Film has the advantages of a much
larger format, color, convenience, and
independence of the wall plug (the ST-4 can be
battery operated in conjunction with a laptop
computer, though). After some use you will find
that film is best for producing sensational color
pictures, and the CCD is best for planets, small
faint objects, and general scientific work such as
variable star monitoring and position
determination. It is for this reason that we
designed the ST-4 to support both efforts, as a
stand-alone tracker, or as an imaging camera.
Our reliance on serial communications slows
down the image update rate, but allows the use
of portable laptop computers which seldom will
accept a PC card.
SBIG ST-4/0489Page 9
SBIGSANTA BARBARA INSTRUMENT GROUP
HOST COMPUTER SOFTWARE OVERVIEW
This section describes the features of the host
software used to interface the ST-4 to either an
IBM PC or compatible or a Macintosh. You will
want to read this section independent of the type
of computer you have, and will also need to read
the section below dealing with the type of
computer you own.
Displaying Images
One of the obvious features of the software is its
ability to display the images taken with the CCD
camera. There is much latitude in the processing
of the images due to the "digital" nature of the
images and the availability of personal computers
with graphics displays.
Adjusting the Contrast
Any image taken by the ST-4 consists of an array
of 192 x 165 pixel values, with each pixel value
having 8 bits of accuracy or in other words values
from 0 (completely dark) to 255 (saturated).
Although any one pixel can have this dynamic
range, typically the interesting aspects of an
image will be constrained to a more limited
range.
There will usually be a background level
associated with the image, which is like a
spatially uniform intensity level, due to dark
current, sky background, or uniform luminosity.
In addition, the bright stars in an image may
saturate, or you may be interested in examining
low-level luminosity (such as nebulosity). For
these reasons, the image's contrast can be
enhanced with the use of two parameters:
Background (sometimes abbreviated as Back) and
Range.
The Background parameter specifies an
intensity level at which any pixel below that
intensity will appear black in the image.
Increasing the Background parameter will have
the effect of masking or hiding the uniform
background, or low-level intensities. The Range
parameter is then used to specify the range of
pixel values above the Background level that will
cause the image to saturate on the display. As an
example setting the Background parameter to 25
and the Range parameter to 50 will cause any
pixels with intensity less than 25 to be displayed
as black, any pixel values between 25 and 75
(25+50) to be displayed with a gray scale, and any
pixels above 75 to be completely white.
You will want to experiment with the settings
of the Background and Range parameters to get a
good feel for how they affect the image. These
parameters do not physically change the image
data, but only affect the way the image is
displayed. We will refer to the processing of the
data to increase the contrast as "stretching the
data".
The software can automatically pick an
appropriate set of values for the Background and
Range parameters using an Auto-Contrast
feature. It does this by noting the number of
pixels at each of the 256 possible intensity levels
(called a Histogram) and setting the Background
and Range parameters based upon that
calculation. This works quite well for images of
extended objects (nebulae, the moon, etc.). For
images of star fields you can try Auto-Contrast,
but will probably get better results by manually
adjusting the Background and Range parameters.
One other feature of the host software in the
image display mode is Presentation Mode, where
the image is centered on the screen (horizontally
and vertically) and the non-image areas of the
screen are blacked-out. This Presentation Mode
is quite handy for taking photographs of the
screen for presentations, etc.
Image Smoothing and Inversion
Another image processing technique available for
the displayed images is image smoothing.
Visually, image smoothing reduces the effects of
noise by smoothing out rapid variations in pixel
brightness. This is accomplished by making each
pixel in the displayed image be a weighted
average of its own pixel value and the values of
its eight neighbors. Finally, the displayed images
can be "Inverted", meaning black areas become
white and vice versa. Visually this is like looking
at a negative, and can produce good results with
images showing subtle nebulosity. Also, it seems
that you can push the contrast harder on Inverted
images without the saturated regions detracting
from the image appearance.
Crosshairs and Photometric Analysis
Due to the linear properties of CCDs and the
digital nature of the image data, photometric
Page 10SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
analysis of CCD images is easily achieved when
compared to techniques used with film. This
section discusses the photometric capabilities of
the ST-4 and its host software.
Scope Factors
Several telescope and calibration factors are
required by the computer software to correctly
calculate star brightnesses and separations. The
telescope focal length in inches and aperture area
in square inches are required, as well as a
calibration factor. The focal length is usually just
the focal length of the telescope, but can be
adjusted for such things as barlows and image
processing techniques such as image zooming.
The calibration factor is the reading that would
be produced from the CCD by a zero magnitude
star focused onto one pixel by a lens with 1
square inch of area, with an integration time of 1
second (with a gain factor of 1). Its value is
typically around 12000, but will vary due to
atmospheric and telescopic transmission, and
CCD device variations. The calibration factor
should also be scaled with the gain (described in
the "Baseline and Gain Parameters" section
below), doubling it for a gain of 2x, tripling it for
a gain of 3x, etc.
Note:
Atmospheric transmission varies with the
elevation of the star, so careful work will
require attention to this detail as well as to
the stellar temperature.
Pixel Coordinates and Intensities
The CCD is configured as illustrated before in
Figure 1. An array of light sensitive detectors,
called pixels, are arranged in an array of 192 by
165 pixels. In this manual and in all other
references we will refer to the 192 pixel wide
dimension as the X-direction and the 165 pixel
tall direction as the Y-direction (see Figure 1).
When an image is displayed on the computer
screen pixel number (1,1) refers to the pixel with
the (x,y) location corresponding to the upper left
hand corner of the screen. X increases to the
right, and Y increases toward the bottom of the
screen. The position of pixel 1,1 on the CCD can
be physically determined by referring to the
placement of the serial number tag on the rear of
the CCD head as shown below in Figure 4.
Pixel 1,1
Figure 4Pixel 1,1 location
The host software allows a crosshair to be
moved across the image, and the coordinates of
the crosshair (in pixels) and the brightness of the
pixel under the crosshair are displayed.
Additionally, a weighted average intensity of the
pixel under the crosshair and its eight neighbors
is also shown. These values of the coordinates,
intensity, and average intensity are the simplest
form of photometry available in the host
software. They are quite handy for setting the
Background and Range parameters and for
determining the image background level (due to
sky background or dark current) and
determining the optimum exposure time.
Magnitudes and Separations
The host software can also measure stellar
magnitudes as well as diffuse magnitudes and
angular separations between objects.
The determination of stellar magnitudes
involves measuring the total light emitted by a
star, or in other words adding the intensity
contributions of all the pixels illuminated by the
star. Star images will rarely be constrained to a
single pixel, hence the requirement for
accumulating the intensity from all illuminated
pixels. In practice, a 5 x 5 box of pixels is used in
determining the magnitude although other size
boxes (3 x 3, 7 x 7, etc.) can sometimes be
specified. Since the magnitude scale is
logarithmic (an increase of 1 magnitude
corresponds to a star which is roughly 2.5 times
as faint) the calculation of stellar magnitude
involves taking the log of the accumulated pixel
intensities. Finally, to accurately measure
SBIG ST-4/0489Page 11
SBIGSANTA BARBARA INSTRUMENT GROUP
magnitudes the background level must be
subtracted. This is done by moving the crosshair
to a dark area of the image and specifying that
region as being typical of the background
intensity for the image. This specification of the
background allows the magnitude calculations to
be independent of effects such as dark current or
sky background. The factors that affect the
calculation of stellar magnitude are: Exposure
time, Aperture area, and Calibration factor.
The diffuse magnitude (also called surface
brightness) of diffuse objects is the magnitude per
square arc-second, and is calculated identically to
the magnitude calculation discussed above
except that the accumulated pixel intensities are
divided by the area of the 5 x 5 box in square
arc-seconds. The factors that affect the
calculation of diffuse magnitude are: Exposure
time, Aperture area, Focal length, and Calibration
Factor.
The host software also allows you to measure
the angular separation between objects in an
image. This is done by moving the crosshair to
the first object, establishing that position as a
reference position, and then moving the crosshair
to the second object. The software then displays
the angular separation and orientation (in
degrees, clockwise of vertical) of the crosshair's
position relative to the fixed reference position.
The calculation of the separation between two
objects is only dependent on the dimensions of
the CCD pixels and the focal length of the
telescope used (which must be accurately
entered). This usually requires experimental
determination using known double stars for
precision. The direction corresponding to eastwest in a setup can be determined by taking an
image of a star, letting the image drift for a few
seconds, and taking another image. The images
can then be co-added and the line between the
two images of the same star delineates the EastWest direction.
Other Image Processing Techniques
The host software uses other image processing
techniques besides contrast enhancement,
smoothing, inversion, and photometric analysis.
These other techniques are discussed in this
section.
Flipping the Image
Some inspection will reveal that the screen image
is flipped about a horizontal axis relative to the
CCD. The horizontal and vertical flip commands
enable a picture to be oriented correctly no
matter what combination of telescope and prisms
is used to form the image. Also, the flip
commands are quite useful for making an image's
orientation match that of published images.
These commands actually modify the image data,
and hence the results of using these commands
are retained if the image is saved after these
commands are applied to an image.
Zooming
The host software allows you to zoom in on an
area in the image by moving a zoom-box over the
image until the zoom-box is positioned at the
desired region where the zoom may be
completed. The zoom-box is quarter sized (48 x
41 pixels), and the pixels within it are then
zoomed to a full 192 x 165 sized image and
interpolation used to fill in the "missing pixels".
The zoomed image can then be used with all the
photometric analysis software, etc., and can also
be saved. Depending on the amount of host
memory available when the zoom is performed,
the zoom either writes over the original image
data (you are warned first) or the original data is
retained and the image can be un-zoomed later
(unless the zoomed data is saved on disk in
which case the original un-zoomed data is
discarded).
Zooming can be quite handy for examining
close binary stars, and small detail, but is no
substitute for higher magnification images since
the zooming process doesn't contain any more
information than the original 192 x 165 pixel
image contained.
Histograms
The host software can also calculate and display
an image's histogram, which is a count of the
number of pixels at each of the 256 possible (0
through 255) intensity levels. The histogram can
be useful for determining the settings of the
Background and Range contrast parameters (like
the Auto-Contrast does), for determining the
dark current or sky background level, and for
determining the optimum exposure time.
Page 12SBIG ST-4/0490
SBIGSANTA BARBARA INSTRUMENT GROUP
Dithering
Although the images from the CCD camera can
have 256 possible intensity levels or gray-scales,
not all computers have the ability to display such
a wide range. Some graphics displays can only
display two colors; black pixels and white pixels
(Hercules adapters and Mac Pluses for example).
Something must be done on these displays or else
images would not look good due to their ultrahigh contrast or lack of gray scale. In these cases
a technique called dithering is used to increase
the number of gray scales available.
Dithering involves using a cell of display
pixels (2 x 2, 3 x 3, etc.) for each image pixel, and
strategically turning on combinations of pixels
within that cell to simulate the gray scale. This
works well because the eye does a good job of
averaging the intensity over the entire cell. The
trade-off is that you require a larger number of
pixels for an image, or must decrease the spatial
resolution of the image to accommodate a fixed
number of pixels.
Dithering can also be used to increase the
gray-scale capabilities of displays with more than
two but a limited number of gray-scales or colors.
For example, a 2 x 2 dithering cell gives 61
different gray scales on a display adapter with
only 16 shades of gray (such as the high-res VGA
mode and some Macintosh II video cards).
Making the Camera Connection
An important aspect of the host software is its
ability to interface to and communicate with the
ST-4 over the serial port. If you have ever dealt
with serial interfaces you should be happy to
know that much effort has been given to make
the host to ST-4 interface as easy as possible
while maintaining a high degree of compatibility
and flexibility.
On power-up, the ST-4 wakes up at 9600
baud, which is quite safe in terms of being able to
establish a reliable communications link with the
camera, but which as also a bit taxing of your
patience in terms of image download times. The
host can however program the ST-4 to
communicate at any baud rate from a lowly 1200
baud to the highest rate of 57.6K baud. The
actual baud rate used will depend on how you
configure the host program and/or the
environment in which the host and ST-4 exist
(mainly the cable length between the host and the
ST-4). The host software can be configured to
talk to the ST-4 at the fixed baud rates of 1200,
9600, 19.2K, and 57.6K or can be configured for
Auto Baud mode.
In the fixed modes, the host software will
program the ST-4 to the selected baud rate and
always attempt to communicate at that rate. In
some circumstances it may be possible to pick too
high a baud rate for reliable communications, and
a lower rate may need to be selected. This is
described further below.
In the Auto Baud mode the host software will
sequentially program the ST-4 for the highest
possible baud rate (57.6K baud), and then test the
communications link, lowering the baud rate as
necessary to establish a link with the camera. The
only times this "auto baud" process will occur is
when the host software is first run or when the
Establish link command is executed under user
control. Another feature of the Auto Baud mode
is that if the host ever looses the communications
link with the ST-4, it will search around (in baud)
until it finds the camera or gives up. In this case
however, it will not try to change the ST-4's baud
rate, but will continue to communicate at the
discovered rate.
We suggest you use the Auto Baud mode,
and let the computer determine the best
communications rate. If you do, you should
always make sure the communications link is
established with the ST-4 prior to downloading
images, by either having the ST-4 connected to
the selected port and powered-up when the host
software is first run, or by explicitly establishing
the communications link using the Establish link
command. Also, if you ever turn off the ST-4 or
loose contact with it you will need to use the
Establish link command to re-program the
highest possible communications rate. If
however you always use the same configuration
(cable length, environment, etc.) then selecting
one of the fixed rates may slightly speed up the
start-up phase of the host software since it won't
have to run the "auto baud" process.
SBIG ST-4/0489Page 13
Loading...
+ 32 hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.